LIBRARY: The use of AI to develop forest fire risk maps

This document analyzes how fire hazard would change in the IPA ADRION area based on two contrasting IPCC climate scenarios up to 2060 and 2100. Using machine learning algorithms such as logistic regression and maximum entropy, the study found significant variations: continental and higher-altitude areas show little change, while Mediterranean regions face greater risk due to reduced rainfall. Precipitation variables proved most influential in driving fire hazard trends. The work led to the definition of a harmonized and transferable procedure for assessing fire hazard now and under climate change, integrating ecological and population vulnerability to map overall fire risk across the region.

Do you want to know more? Click here to download the “Deliverable 1.3.1 Standardisation of procedures and data harmonisation for wildfire risk mapping”